The objective of this study was to define a simple method of measuring thoracic aortic pulse-wave velocity (PWV) using only an aortic length regression equation based on the subject’s height and two pulsed Doppler recordings of the aorta.
The thoracic aortas of 80 subjects (age range, 0-20 years) were measured retrospectively by direct echocardiographic visualization. A simple linear regression equation for thoracic aortic length on the basis of height was derived. PWV was defined as the thoracic aortic length, derived from the subject’s height, divided by pulse transit time. Pulse transit time was defined as the difference in the time of onset of two pulsed Doppler recordings placed at the level of the aortic valve leaflet tips and the diaphragm. Normative data were retrospectively defined in 206 children.
Thoracic aortic length was linearly related to subject height by the equation thoracic aortic length (cm) = 1.7 cm + 0.1 (height [cm]) ( R 2 = 0.98, P < .0001). Thoracic aortic PWV was independent of age (median, 3.04 m/s).
Thoracic aortic PWV can be simply calculated from a routine echocardiogram, it is constant throughout childhood, and it may improve the assessment of left ventricular load.
Aortic pulse-wave velocity (PWV; aortic length/pulse transit time) is the speed at which an arterial wave propagates through a segment of the aorta. Central aortic PWV is a marker of aortic wall stiffness that has been shown in adult studies to predict cardiovascular morbidity and mortality in the elderly as well as in patients with end-stage renal disease and hypertension. A direct correlation between increased PWV in childhood and mortality has yet to be established. Several childhood conditions, including coarctation of the aorta, tetralogy of Fallot, and Marfan syndrome, have documented elevations in aortic PWV. It is also well established that the process of atherogenesis and the development of cardiovascular risk factors begins in childhood. Several of these known cardiovascular risk factors have been correlated with aortic wall stiffness in children, including obesity, physical inactivity, hypercholesterolemia, insulin resistance, and diabetes. However, it is unknown whether modifying vascular changes, including PWV, in these childhood conditions decreases overall cardiovascular risk. A simple method of measuring aortic PWV in children would promote future research in this area.
Central aortic stiffness has been evaluated in different arterial locations by multiple methods, each having limitations. Catheterization methods that can directly measure aortic length and pulse pressure are too invasive for routine use. More practical, noninvasive methods include pressure tonometry, magnetic resonance imaging (MRI), and echocardiographic methods.
The carotid-femoral method by pressure tonometry is the “gold standard” and the most widely used method of obtaining central aortic PWV, as recommended by the first consensus conference on arterial stiffness in 2000. The carotid-femoral method measures the average PWV of the aorta from the descending aorta to the femoral artery using pressure waveforms obtained from tonometry. Its correlation with morbidity and mortality has been well established. However, the method makes two significant assumptions that may increase measurement error: (1) aortic distance can be approximated well by body surface measurements, and (2) aortic flow onset at the carotid artery is a good surrogate marker for flow onset in the descending aorta at a location equidistant from the aortic valve. Finally, a specific limitation of this method in our population is that it can be challenging to obtain measurable pressure waveforms in young, uncooperative children. The carotid arteries of children may be obscured by their short, thick necks, and they are often not able to remain still long enough to record pressure waveforms. To overcome some of these limitations, MRI and echocardiographic methods have been established. These methods determine PWV by measuring distance by direct arch visualization and transit time by using arterial flow waves.
MRI measures of aortic PWV have been described. MRI provides direct aortic length measurement and allows for subsegmental PWV analysis of the aorta. Limitations of this method include the extensive amount of time and the financial expense involved in performing the procedure. In addition, young children may require general anesthesia to acquire these images.
An alternative method of measuring the proximal aortic PWV in children using echocardiography has been described. Distance from the aortic valve to the proximal descending aorta is directly measured. Transit time is measured using two pulsed Doppler recordings obtained from this view, one in the aortic root and one in the descending aorta. The authors describe that difficult acoustic windows often limit arch visualization and measurement. In addition, the location of the Doppler measurement in the descending aorta is not standardized. Different segments of the aorta have different PWVs. Therefore, ending the PWV measurement at various locations may change the PWV result.
The main limitation of the noninvasive methods of obtaining central aortic PWV is the difficulty involved in accurately measuring the aortic length. The aim of this study was to define a simple and practical method of measuring PWV in a defined length of the proximal aorta that is easily obtainable in all patients without the need for adequate acoustic windows to directly measure the arch. To do so, the first objective was to derive a linear regression equation for thoracic aortic length on the basis of patient height. A regression equation obviates the need for aortic length measurements to be performed in each individual. The second objective was to define a new simplified measure of thoracic aortic PWV using the height-based thoracic aortic length and two pulsed Doppler recordings of the aorta. The third objective was to define normative data for thoracic aortic PWV in healthy normotensive children on the basis of this method.
All subjects were derived from a retrospective review of our echocardiogram archive from January 2006 to June 2007. At the time of echocardiography, some children were sedated with chloral hydrate. Images were acquired using a Siemens Acuson Sequoia (Siemens Medical Solutions USA, Inc, Mountain View, CA) and stored in a digital archive using Joint Photographic Experts Group compression. Institutional review board approval was obtained. The study was exempt from the requirement for informed consent.
Thoracic Aortic Length Regression Equation
To derive a regression equation relating thoracic aortic length to patient height, the echocardiograms of 80 patients were reviewed. Subjects were identified by searching the database for the following study indications: murmur, ventricular septal defect, atrial septal defect, syncope, arrhythmia, or adverse effect of chemotherapy. Subjects were included if age was 0 to 20 years and height was 30 to 190 cm. Five patients of each sex at 20-cm height intervals were included. Studies required a subcostal sagittal view clearly depicting the entire thoracic aorta from the aortic valve to the level of the diaphragm. Subjects were excluded if they had aortic arch anomalies or known genetic syndromes.
Electronic calipers were used to measure thoracic aortic length, from the aortic valve leaflet tips to the level of the diaphragm in the subcostal sagittal view ( Figure 1 ). Measurements were performed independently by two investigators. Results were averaged. Simple linear regression was used to establish the relationship between thoracic aortic length and the subject’s height.
Thoracic Aortic PWV Calculation
Thoracic aortic length was calculated from the regression equation on the basis of height.
Transit time was defined as the difference in the time of onset of flow at the diaphragm and the aortic valve, measured by pulsed Doppler using the electrocardiogram as a time reference. Pulsed Doppler recordings were performed at a sweep speed of 100 m/s, with subjects in normal sinus rhythm.
Pulsed Doppler recordings of the ascending aorta were obtained from the apical view with the sample volume placed at the level of the valve leaflet tips. Pulsed Doppler recordings of the descending aorta were obtained from the subcostal sagittal view with the sample volume placed in the center of the aorta at the level of the diaphragm ( Figure 2 ). The time from the onset of the QRS complex to the foot of the aortic waveform was measured at each location ( Figure 3 ).
PWV was calculated as follows: PWV (m/s) = height-based aortic length (cm)/(100 × transit time [s]).
Interobserver correlation of transit time measured by two observers was performed in 105 subjects.
Normative Data for Thoracic Aortic PWV
To define normative data for thoracic aortic PWV in children aged 0 to 15 years, echocardiograms of 206 healthy normotensive children were retrospectively reviewed. Subjects were identified by searching the database for the following study indications: murmur, chest pain, small ventricular septal defect, small atrial septal defect, patent foramen ovale, and trivial patent ductus arteriosus. Subjects with aortic anomalies and hemodynamically significant cardiac anomalies, including evidence of volume load secondary to their lesions, were excluded.
Thoracic aortic PWV was calculated as described above. The results of 2 or 3 pulsed Doppler recordings at each location were averaged. Because this was a retrospective study, the location of the pulsed Doppler sample volume was not always placed exactly at the diaphragm. The exact location of the pulsed Doppler sample was recorded in reference to the distance above or below the diaphragm. The aortic length estimate was adjusted accordingly.
A simple linear regression equation between thoracic aortic length and height was determined. Goodness of fit to the line was assessed by a coefficient of determination. Interobserver reproducibility of measurements was assessed by a correlation coefficient.
Because of the positive skewness of thoracic aortic PWV, the natural logarithm of this variable was used to meet the underlying assumption of our model. Multiple linear regression was carried out to test the effects of sex and age on the thoracic aortic PWV.
Thoracic aortic length was related to height by a simple linear regression equation: thoracic aortic length (cm) = 1.7 cm + 0.10 (height [cm]) ( R 2 = 0.98, P < .0001; Figure 4 ). Interobserver reproducibility of measurements (by two observers) for thoracic aortic length was excellent ( R = 0.98, P < .0001). Intraclass correlation for the transit time measurement was good (intraclass correlation coefficient, 0.91).